© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 19945
    Prev
    Next
    INDEX
    Explanations

    words related to technical topics and instructions

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    36
    +0.13
    0.6%
    286
    +0.10
    0.5%
    316
    +0.09
    0.4%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    36
    +0.13
    0.03
    395
    +0.10
    0.02
    316
    +0.09
    0.02
    Negative Logits
    enance
    -0.93
    antry
    -0.77
    anism
    -0.71
    hyde
    -0.70
    peed
    -0.70
    oyer
    -0.64
    independence
    -0.64
    ionics
    -0.64
    information
    -0.63
    iliation
    -0.63
    POSITIVE LOGITS
     few
    1.37
     handful
    1.30
     bunch
    1.24
     dozen
    1.22
     lot
    1.20
     plethora
    1.16
     couple
    1.14
     slew
    1.04
     myriad
    1.03
     multitude
    1.02
    Activations Density 0.049%

    No Known Activations